Page - 3 - in Applied Interdisciplinary Theory in Health Informatics - Knowledge Base for Practitioners

Box 1. Schneiderman’s checklist of eight user interface feature associated with high usability [2] 1 Strive for consistency 2 Enable frequent users to use shortcuts 3 Offer informative feedback 4 Design dialog to yield closure 5 Offer simple error handling 6 Permit easy reversal of actions 7 Support internal locus of control 8 Reduce short-term memory load 5. Evaluation frameworks: a framework or checklist that specifies aspects of a project that can be evaluated to determine if it is successful. An influential example in health informatics is Kidholm’s Model for ASsessment of Telemedicine applications, the MAST framework [5]. In this book, process models and evaluation frameworks are considered to be of less importance than determinant frameworks, classic theories and health informatics theories, so I will confine our discussion to these three categories, referring to them henceforth as “theories”. It is no exaggeration to state that the editors and most authors of chapters in this book would agree that the identification, testing and use of theories is crucial to the future maturation of health informatics as a recognized profession, for at least five reasons. First, no one would argue that we currently know how to produce usable, effective clinical systems every time – indeed, it seems that sometimes success in clinical informatics is the exception rather than the rule [6]. So, we need predictive theories to make health information systems better: that is, more usable, better accepted, more accurate, more clinically and cost effective, and readily transferable to other settings. Second, we need theories to help us build an evidence and scientific basis for our discipline, to help it evolve from a craft - based on anecdote, apprenticeship and learning from mistakes - to a professional engineering discipline [7] similar to, for example, the development of aeronautical engineering. Box 2 describes an example from aeronautical engineering of how formulating and testing a theory became a key method both to enhance aircraft safety and to promote the emergence of a professional discipline. Third, we need theories (and an understanding about which theory to use, and when) to teach our students and practitioners. Fourth, we need theories to guide organisations procuring systems, so that they can distinguish between theory-based systems that are likely to be effective from atheoretical systems which are less likely to help. Finally, we need a list of tested theories (both useful and useless theories) to help decide rationally whether to carry out a full evaluation of a clinical system following an update or not, according to whether the components that were theory-based are still included. There is an analogy here with medical devices regulation [9]: a previously approved cardiac catheter does not need further testing and regulatory approval if the changes are minor, but it does if the changes are “material”. In our case, we could be confident if a lifestyle app, for example, is altered in a minor way, but not if theory-based behavior change features are removed. J.C.Wyatt /TheNeed forTheory to InformClinical InformationSystems 3